Class: Feed

Inherits:
Object
  • Object
show all
Defined in:
lib/kindle-feeds.rb

Class Method Summary collapse

Class Method Details

.auto_discover_and_subscribe(url) ⇒ Object 



166
167
168
169
170
171
172
173
174
175
176
177
 
# File 'lib/kindle-feeds.rb', line 166

def self.auto_discover_and_subscribe(url)
  uri = URI.parse(url)
  feed_url = Autodiscovery.new(fetch(url)).discover
  if feed_url
    feed_url = uri.merge(feed_url).to_s
    puts "Found feed: #{feed_url}" 
    return feed_url
  else
    puts "Can't find feed for #{url}" 
    return nil
  end
end

.create_feed(xml, feed_url) ⇒ Object 



93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
 
# File 'lib/kindle-feeds.rb', line 93

def self.create_feed(xml, feed_url)
  begin
    feed = FeedNormalizer::FeedNormalizer.parse(xml)
    return nil unless feed.is_a?(FeedNormalizer::Feed)
  rescue
    puts "Error trying to parse feed: #{feed_url}"
    puts $!
    puts
    return nil
  end
  # clean up entries:
  ic = Iconv.new('ISO-8859-1//TRANSLIT', 'utf-8') 
  ic2 = Iconv.new('ISO-8859-1//IGNORE', 'utf-8') 
  #ic = Iconv.new('ASCII//TRANSLIT', 'utf-8') 
  coder = HTMLEntities.new
  puts "#{feed.entries.size} entries downloaded."
  puts
  feed.entries.each do |e|
    e.title = coder.decode(e.title)
    e.content = coder.decode(e.content)

    begin
      e.title = ic.iconv(e.title) 
    rescue
      e.title = ic2.iconv(e.title) 
    end
    begin
      e.content = ic.iconv(e.content) 
    rescue
      e.content = ic2.iconv(e.content) 
    end
    doc = Hpricot(e.content)
    doc.search('h1, h2, h3') do |h|
      h.swap("<h4>#{h.inner_text}</h4>")
    end
    doc.search('//font') do |font|
      font.swap(font.inner_text)
    end
    doc.search('//img').remove
    doc.search('svg, object, embed').remove
    doc.search('script').remove
    e.content = doc.to_s
  end
  return feed
end

.fetch(url) ⇒ Object

a simple wrapper over open-uri call. Easier to mock in testing.



180
181
182
183
184
185
186
187
188
 
# File 'lib/kindle-feeds.rb', line 180

def self.fetch(url)
  begin
    open(url).read
  rescue Timeout::Error 
    puts "-> attempt to fetch #{url} timed out"
  rescue Exception => e
    puts "-> error trying to fetch #{url}: #{$!}"
  end
end

.subscribe(feed_url) ⇒ Object

try to repair the URL if possible



139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
 
# File 'lib/kindle-feeds.rb', line 139

def self.subscribe(feed_url) # try to repair the URL if possible
  unless feed_url =~ /^http:\/\//
    feed_url = "http://" + feed_url
  end
  puts "Downloading #{feed_url}"
  begin
    xml = fetch(feed_url)
  rescue SocketError
    puts "Error trying to load page at #{feed_url}"
    return
  end
  if xml.nil?
    puts "Can't find any resource at #{feed_url}"
    return
  end
  feed = Feed.create_feed( xml, feed_url.strip )
  if feed.nil?
    puts "#{feed_url}: Attempting autodiscovery..."
    feed_url = auto_discover_and_subscribe(feed_url)
    if feed_url
      xml = fetch(feed_url)
      feed = Feed.create_feed( xml, feed_url.strip )
    end
  end
  feed
end